mlsanityData to 7 Oct 2026

Challenges and Practices of Deep Learning Model Reengineering: A Case Study on Computer Vision

arXiv 2303.07476 ↗First posted 13 Mar 2023cs.SE cs.AIcited by 11

Not on the map. Why?

Abstract (arXiv)

Many engineering organizations are reimplementing and extending deep neural networks from the research community. We describe this process as deep learning model reengineering. Deep learning model reengineering - reusing, reproducing, adapting, and enhancing state-of-the-art deep learning approaches - is challenging for reasons including under-documented reference models, changing requirements, and the cost of implementation and testing. In addition, individual engineers may lack expertise in software engineering, yet teams must apply knowledge of software engineering and deep learning to succeed. Prior work has examined on DL systems from a "product" view, examining defects from projects regardless of the engineers' purpose. Our study is focused on reengineering activities from a "process" view, and focuses on engineers specifically engaged in the reengineering process. Our goal is to understand the characteristics and challenges of deep learning model reengineering. We conducted a case study of this phenomenon, focusing on the context of computer vision. Our results draw from two data sources: defects reported in open-source reeengineering projects, and interviews conducted with open-source project contributors and the leaders of a reengineering team. Our results describe how deep learning-based computer vision techniques are reengineered, analyze the distribution of defects in this process, and discuss challenges and practices. Integrating our quantitative and qualitative data, we proposed a novel reengineering workflow. Our findings inform several future directions, including: measuring additional unknown aspects of model reengineering; standardizing engineering practices to facilitate reengineering; and developing tools to support model reengineering and model reuse.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Builds on 56 of its 184 references that are in the map. First posted on arXiv 13 Mar 2023.

Builds on

56 of 184 references.

References · 56

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  4. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  5. 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,829
  6. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,530
  7. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  8. 2021Masked-attention Mask Transformer for Universal Image Segmentationcited by 2,338
  9. 2017Accurate, Large Minibatch SGD: Training ImageNet in 1 Hourcited by 1,479
  10. 2016Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translationcited by 1,471
  11. 2017Towards A Rigorous Science of Interpretable Machine Learningcited by 1,400
  12. 2019Dota 2 with Large Scale Deep Reinforcement Learningcited by 1,084
38 more
  1. 2018Model Cards for Model Reportingcited by 1,068
  2. 2021Florence: A New Foundation Model for Computer Visioncited by 802
  3. 2014Deep Learning in Neural Networks: An Overviewcited by 765
  4. 2015An Introduction to Convolutional Neural Networkscited by 510
  5. 2020A Survey of Convolutional Neural Networks: Analysis, Applications, and Prospectscited by 396
  6. 2022Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Modelscited by 379
  7. 2020Improving Reproducibility in Machine Learning Research (A Report from the NeurIPS 2019 Reproducibility Program)cited by 318
  8. 2017DeepXplore: Automated Whitebox Testing of Deep Learning Systemscited by 307
  9. 2019Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentationcited by 258
  10. 2017DeepTest: Automated Testing of Deep-Neural-Network-driven Autonomous Carscited by 249
  11. 2019Machine Learning Testing: Survey, Landscapes and Horizonscited by 213
  12. 2021LogME: Practical Assessment of Pre-trained Models for Transfer Learningcited by 168
  13. 2018DeepGauge: Multi-Granularity Testing Criteria for Deep Learning Systemscited by 144
  14. 2019Taxonomy of Real Faults in Deep Learning Systemscited by 90
  15. 2019A Comprehensive Study on Deep Learning Bug Characteristicscited by 65
  16. 2018DLFuzz: Differential Fuzzing Testing of Deep Learning Systemscited by 64
  17. 2021Collaboration Challenges in Building ML-Enabled Systems: Communication, Documentation, Engineering, and Processcited by 56
  18. 2018On Testing Machine Learning Programscited by 48
  19. 2023An Empirical Study of Pre-Trained Model Reuse in the Hugging Face Deep Learning Model Registrycited by 45
  20. 2020A Survey of Methods for Low-Power Deep Learning and Computer Visioncited by 37
  21. 2021DeepLocalize: Fault Localization for Deep Neural Networkscited by 37
  22. 2022On the Power of Foundation Modelscited by 37
  23. 2020Repairing Deep Neural Networks: Fix Patterns and Challengescited by 33
  24. 2020Adoption and Effects of Software Engineering Best Practices in Machine Learningcited by 33
  25. 2022NNSmith: Generating Diverse and Valid Test Cases for Deep Learning Compilerscited by 29
  26. 2022Toward Understanding Deep Learning Framework Bugscited by 26
  27. 2018What's in a GitHub Star? Understanding Repository Starring Practices in a Social Coding Platformcited by 25
  28. 2022Towards Training Reproducible Deep Learning Modelscited by 23
  29. 2021Production Machine Learning Pipelines: Empirical Analysis and Optimization Opportunitiescited by 21
  30. 2020On the Replicability and Reproducibility of Deep Learning in Software Engineeringcited by 16
  31. 2021Decomposing Convolutional Neural Networks into Reusable and Replaceable Modulescited by 14
  32. 2021Automated Evolutionary Approach for the Design of Composite Machine Learning Pipelinescited by 13
  33. 2019Machine Learning Software Engineering in Practice: An Industrial Case Studycited by 12
  34. 2021Design Smells in Deep Learning Programs: An Empirical Studycited by 10
  35. 2019Analysis and Detection of Information Types of Open Source Software Issue Discussionscited by 9
  36. 2020Deep Learning & Software Engineering: State of Research and Future Directionscited by 9
  37. 2021An Experience Report on Machine Learning Reproducibility: Guidance for Practitioners and TensorFlow Model Garden Contributorscited by 6
  38. 2022Towards a Change Taxonomy for Machine Learning Systemscited by 6

50 most cited of 56 All 56 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 3; 2024: 3; 2025: 4; 2026: 1.

11 papers.

  1. 2024PeaTMOSS: A Dataset and Initial Analysis of Pre-Trained Models in Open-Source Softwarecited by 9
  2. 2024What do we know about Hugging Face? A systematic literature review and quantitative validation of qualitative claimscited by 8
  3. 2025An Empirical Analysis of Machine Learning Model and Dataset Documentation, Supply Chain, and Licensing Challenges on Hugging Facecited by 6
  4. 2023Test & Evaluation Best Practices for Machine Learning-Enabled Systemscited by 3
  5. 2025Improving the Reproducibility of Deep Learning Software: An Initial Investigation through a Case Study Analysiscited by 3
  6. 2023PeaTMOSS: Mining Pre-Trained Models in Open-Source Softwarecited by 1
  7. 2024A Partial Replication of MaskFormer in TensorFlow on TPUs for the TensorFlow Model Gardencited by 1
  8. 2025How Do Model Export Formats Impact the Development of ML-Enabled Systems? A Case Study on Model Integrationcited by 1
  9. 2026An Empirical Investigation of Pre-Trained Deep Learning Model Reuse in the Scientific Processcited by 1
  10. 2023Analysis of Failures and Risks in Deep Learning Model Converters: A Case Study in the ONNX Ecosystemcited by 0
  11. 2025NeMo: A Neuron-Level Modularizing-While-Training Approach for Decomposing DNN Modelscited by 0